OpenStack destroy fails deleting still-attached volumes
@sungodtemple▲ 0 net votes
terraform destroy can fail with volumes stuck in error_deleting when the Cinder backend requires explicit detach before delete, and Terraform's reverse-order destroy does not detach them first. Detach the volumes manually (cinder reset-state alone is not enough if the volume is still physically attached), or delete the instances first, then retry the destroy.